Methods and Considerations for Hypothesis-free Signal Detection Studies Accommodating Various Types of Medications, Populations, and Regions

Background: Hypothesis-free signal detection is crucial in the initial phase of drug safety surveillance as it generates hypotheses of potential drug-event pairs for further evaluation. The conduct of hypothesis-free analyses should accommodate different scenarios, including medication types, populations, and geographic regions.

Objectives: To exchange opinions on considerations of conducting signal detection studies across different scenarios. To obtain consensus on methods accommodating various types of medications, populations, geographic regions, and other settings. The target audience are all ISPE members who are interested in signal detection methods.

Description: Miyuki Hsieh (10 mins) will give an overview of the symposium, focusing on available databases and methods for hypothesis-free signal detection. She will point out challenges of conducting signal detection analyses in different settings. Five speakers from different continents will share their experiences on signal detection studies and highlight the consideration on the methods accommodating different perspectives, including:
(1) Signal detection for drug vs vaccine (Andrew Bate, UK; 12mins)
(2) Signal detection for small molecule drugs vs biologics (Nicole Pratt, Australia; 12mins)
(3) Signal detection for single drug vs drug-drug interaction (Nam-Kyong Choi, Korea; 12mins)
(4) Signal detection for general population vs pregnant women (Judith Maro, USA; 12mins)
(5) Single country vs. cross-country signal detection (Lars Christian Lund, Denmark; 12 mins)

Professor Edward Lai from Taiwan and Professor Darren Toh from the US will moderate the panel discussion (20 mins). Select discussion topics are listed below. A publicly available interactive QA tool, Slido, will be utilized in the panel discussion to collect and prioritize the questions from the audiences.

(1) Challenges of signal detection studies in different scenarios and potential solutions.
(2) Consensus on methods consideration across different scenarios.
(3) How to interpret results from signal detection studies and to establish strategies to improve drug safety monitoring.
(4) Possibility for collaborative studies across continents to improve the statistical power and generalizability of study results, and the efficiency of signal detection.

At the end of the symposium, we will summarize potential challenges, considerations and solutions on signal detection methods and build consensus for methods accommodating different settings.